OpenBIOS February 2010

openbios@openbios.org

12 participants
22 discussions

[commit] r683 - trunk/openbios-devel/config/examples
by repository service 23 Feb '10

23 Feb '10

Author: agraf Date: Mon Feb 22 19:52:34 2010 New Revision: 683 URL: http://tracker.coreboot.org/trac/openbios/changeset/683 Log: PPC: Disable host binary build On PPC the host binary build always fails for me. Since I guess I'm the only one actually building OpenBIOS ppc on ppc and I don't really care that much about debugging Forth code in my Linux environment, I guess we can just disable building it by default. Signed-off-by: Alexander Graf <agraf(a)suse.de> Modified: trunk/openbios-devel/config/examples/ppc_config.xml Modified: trunk/openbios-devel/config/examples/ppc_config.xml ============================================================================== --- trunk/openbios-devel/config/examples/ppc_config.xml Mon Feb 22 19:52:32 2010 (r682) +++ trunk/openbios-devel/config/examples/ppc_config.xml Mon Feb 22 19:52:34 2010 (r683) @@ -10,7 +10,7 @@ <option name="CONFIG_BIG_ENDIAN" type="boolean" value="true"/>  - <option name="CONFIG_HOST_UNIX" type="boolean" value="true"/> + <option name="CONFIG_HOST_UNIX" type="boolean" value="false"/> <option name="CONFIG_UNIX_QT" type="boolean" value="false"/> <option name="CONFIG_PLUGINS" type="boolean" value="false"/>

1 0

[commit] r682 - in trunk/openbios-devel: arch/ppc/qemu drivers
by repository service 23 Feb '10

23 Feb '10

Author: agraf Date: Mon Feb 22 19:52:32 2010 New Revision: 682 URL: http://tracker.coreboot.org/trac/openbios/changeset/682 Log: PPC: Set Uninorth interrupt numbers to new Qemu values We changed the uninorth interrupt mapping to better reflect real hardware. OpenBIOS obviously needs to know about that. Signed-off-by: Alexander Graf <agraf(a)suse.de> Modified: trunk/openbios-devel/arch/ppc/qemu/init.c trunk/openbios-devel/drivers/pci_database.c Modified: trunk/openbios-devel/arch/ppc/qemu/init.c ============================================================================== --- trunk/openbios-devel/arch/ppc/qemu/init.c Mon Feb 22 19:52:30 2010 (r681) +++ trunk/openbios-devel/arch/ppc/qemu/init.c Mon Feb 22 19:52:32 2010 (r682) @@ -93,9 +93,9 @@ [ARCH_MAC99] = { "MAC99", PCI_VENDOR_ID_APPLE, PCI_DEVICE_ID_APPLE_UNI_N_PCI, 0xf2800000, 0xf2c00000, - 0xf2000000, 0x02000000, 0x80000000, 0x10000000, - 0xf2000000, 0x00800000, 0x00000000, 0x01000000, - { 8, 9, 10, 11 } + 0xf2000000, 0x02000000, 0x80000000, 0x10000000, + 0xf2000000, 0x00800000, 0x00000000, 0x01000000, + { 0x1b, 0x1c, 0x1d, 0x1e } }, [ARCH_MAC99_U3] = { "MAC99_U3", PCI_VENDOR_ID_APPLE, PCI_DEVICE_ID_APPLE_U3_AGP, Modified: trunk/openbios-devel/drivers/pci_database.c ============================================================================== --- trunk/openbios-devel/drivers/pci_database.c Mon Feb 22 19:52:30 2010 (r681) +++ trunk/openbios-devel/drivers/pci_database.c Mon Feb 22 19:52:32 2010 (r682) @@ -289,7 +289,7 @@ PCI_VENDOR_ID_APPLE, PCI_DEVICE_ID_APPLE_UNI_N_AGP, NULL, "pci", "AAPL,UniNorth", "uni-north\0", 3, 2, 1, - NULL, NULL + host_config_cb, NULL, }, { PCI_VENDOR_ID_APPLE, PCI_DEVICE_ID_APPLE_UNI_N_PCI, NULL,

1 0

[commit] r681 - in trunk/openbios-devel: arch/ppc/qemu drivers include/openbios
by repository service 23 Feb '10

23 Feb '10

Author: agraf Date: Mon Feb 22 19:52:30 2010 New Revision: 681 URL: http://tracker.coreboot.org/trac/openbios/changeset/681 Log: PPC: Add U3 based Mac machine Linux on PPC64 knows only so many chipsets. One of the is the U3. So we can use that when we're emulating a PPC64 machine, making Linux happy. Signed-off-by: Alexander Graf <agraf(a)suse.de> Modified: trunk/openbios-devel/arch/ppc/qemu/init.c trunk/openbios-devel/drivers/pci_database.c trunk/openbios-devel/include/openbios/pci.h Modified: trunk/openbios-devel/arch/ppc/qemu/init.c ============================================================================== --- trunk/openbios-devel/arch/ppc/qemu/init.c Mon Feb 22 19:52:28 2010 (r680) +++ trunk/openbios-devel/arch/ppc/qemu/init.c Mon Feb 22 19:52:30 2010 (r681) @@ -63,6 +63,7 @@ ARCH_PREP = 0, ARCH_MAC99, ARCH_HEATHROW, + ARCH_MAC99_U3, }; int is_apple(void) @@ -77,7 +78,8 @@ int is_newworld(void) { - return machine_id == ARCH_MAC99; + return (machine_id == ARCH_MAC99) || + (machine_id == ARCH_MAC99_U3); } static const pci_arch_t known_arch[] = { @@ -95,6 +97,13 @@ 0xf2000000, 0x00800000, 0x00000000, 0x01000000, { 8, 9, 10, 11 } }, + [ARCH_MAC99_U3] = { "MAC99_U3", PCI_VENDOR_ID_APPLE, + PCI_DEVICE_ID_APPLE_U3_AGP, + 0xf0800000, 0xf0c00000, + 0xf0000000, 0x02000000, 0x80000000, 0x10000000, + 0xf2000000, 0x00800000, 0x00000000, 0x01000000, + { 0x1b, 0x1c, 0x1d, 0x1e } + }, [ARCH_HEATHROW] = { "HEATHROW", PCI_VENDOR_ID_MOTOROLA, PCI_DEVICE_ID_MOTOROLA_MPC106, 0xfec00000, 0xfee00000, @@ -557,6 +566,7 @@ break; case ARCH_MAC99: + case ARCH_MAC99_U3: case ARCH_PREP: default: @@ -597,17 +607,6 @@ push_str("system-id"); fword("property"); - /* pci info */ - - if (machine_id == ARCH_MAC99) { - push_str("/pci"); - fword("find-device"); - push_str("u3-agp"); - fword("encode-string"); - push_str("compatible"); - fword("property"); - } - /* memory info */ push_str("/memory"); Modified: trunk/openbios-devel/drivers/pci_database.c ============================================================================== --- trunk/openbios-devel/drivers/pci_database.c Mon Feb 22 19:52:28 2010 (r680) +++ trunk/openbios-devel/drivers/pci_database.c Mon Feb 22 19:52:30 2010 (r681) @@ -280,6 +280,12 @@ static const pci_dev_t hbrg_devices[] = { { + PCI_VENDOR_ID_APPLE, PCI_DEVICE_ID_APPLE_U3_AGP, NULL, + "pci", "AAPL,UniNorth", "u3-agp\0", + 3, 2, 1, + host_config_cb, NULL, + }, + { PCI_VENDOR_ID_APPLE, PCI_DEVICE_ID_APPLE_UNI_N_AGP, NULL, "pci", "AAPL,UniNorth", "uni-north\0", 3, 2, 1, @@ -1084,7 +1090,7 @@ { PCI_VENDOR_ID_APPLE, PCI_DEVICE_ID_APPLE_UNI_N_KEYL, "mac-io", "mac-io", "AAPL,Keylargo", "Keylargo\0", - 1, 1, 2, + 1, 1, 1, &macio_keylargo_config_cb, NULL, }, { Modified: trunk/openbios-devel/include/openbios/pci.h ============================================================================== --- trunk/openbios-devel/include/openbios/pci.h Mon Feb 22 19:52:28 2010 (r680) +++ trunk/openbios-devel/include/openbios/pci.h Mon Feb 22 19:52:30 2010 (r681) @@ -187,6 +187,7 @@ #define PCI_DEVICE_ID_APPLE_UNI_N_PCI 0x001f #define PCI_DEVICE_ID_APPLE_UNI_N_AGP 0x0020 #define PCI_DEVICE_ID_APPLE_UNI_N_KEYL 0x0022 +#define PCI_DEVICE_ID_APPLE_U3_AGP 0x004b #define PCI_VENDOR_ID_SUN 0x108e #define PCI_DEVICE_ID_SUN_EBUS 0x1000

1 0

[commit] r680 - trunk/openbios-devel/drivers
by repository service 23 Feb '10

23 Feb '10

Author: agraf Date: Mon Feb 22 19:52:28 2010 New Revision: 680 URL: http://tracker.coreboot.org/trac/openbios/changeset/680 Log: PPC: Fix interrupt numbers We changed several interrupt numbers in Qemu to better reflect real world hardware. Also, since we're now using proper interrupt maps, we need to make sure we specify interrupts in the way the respective map requires it. Signed-off-by: Alexander Graf <agraf(a)suse.de> Modified: trunk/openbios-devel/drivers/cuda.c trunk/openbios-devel/drivers/ide.c Modified: trunk/openbios-devel/drivers/cuda.c ============================================================================== --- trunk/openbios-devel/drivers/cuda.c Mon Feb 22 19:52:26 2010 (r679) +++ trunk/openbios-devel/drivers/cuda.c Mon Feb 22 19:52:28 2010 (r680) @@ -197,7 +197,12 @@ set_property(ph, "reg", (char *)&props, sizeof(props)); - set_int_property(ph, "interrupts", 0x12); + /* on newworld machines the cuda is on interrupt 0x19 */ + + props[0] = 0x19; + props[1] = 0; + NEWWORLD(set_property(ph, "interrupts", (char *)props, sizeof(props))); + NEWWORLD(set_int_property(ph, "#interrupt-cells", 2)); /* we emulate an oldworld hardware, so we must use * non-standard oldworld property (needed by linux 2.6.18) Modified: trunk/openbios-devel/drivers/ide.c ============================================================================== --- trunk/openbios-devel/drivers/ide.c Mon Feb 22 19:52:26 2010 (r679) +++ trunk/openbios-devel/drivers/ide.c Mon Feb 22 19:52:28 2010 (r680) @@ -1421,9 +1421,11 @@ dnode = find_dev(nodebuff); +#ifndef CONFIG_PPC props[0]=14; props[1]=0; set_property(dnode, "interrupts", (char *)&props, 2*sizeof(cell)); +#endif props[0] = __cpu_to_be32(chan->io_regs[0]); props[1] = __cpu_to_be32(1); props[2] = __cpu_to_be32(8); @@ -1578,10 +1580,31 @@ OLDWORLD(set_property(dnode, "AAPL,pio-timing", (char *)&props, 8*sizeof(cell))); - props[0] = 0x0000000d; - props[1] = 0x00000000; + /* The first interrupt entry is the ide interrupt, the second + the dbdma interrupt */ + switch (current_channel) { + case 1: + props[0] = 0x0000000d; + props[2] = 0x00000002; + break; + case 2: + props[0] = 0x0000000e; + props[2] = 0x00000003; + break; + case 3: + props[0] = 0x0000000f; + props[2] = 0x00000004; + break; + default: + props[0] = 0x00000000; + props[2] = 0x00000000; + break; + } + props[1] = 0x00000000; /* XXX level triggered on real hw */ + props[3] = 0x00000000; set_property(dnode, "interrupts", - (char *)&props, 2*sizeof(cell)); + (char *)&props, 4*sizeof(cell)); + set_int_property(dnode, "#interrupt-cells", 2); OLDWORLD(set_property(dnode, "AAPL,interrupts", (char *)&props, 2*sizeof(cell)));

1 0

[commit] r679 - in trunk/openbios-devel: drivers include/openbios
by repository service 23 Feb '10

23 Feb '10

Author: agraf Date: Mon Feb 22 19:52:26 2010 New Revision: 679 URL: http://tracker.coreboot.org/trac/openbios/changeset/679 Log: PPC: Create interrupt map for PCI There is a draft spec that defines how interrupts are supposed to be mapped from one bus to another, down to the PIC and CPU interrupt line. So far we don't implement that spec. But PPC64 Linux requires us to for PCI devices. So let's create a map here. Draft: http://playground.sun.com/1275/practice/imap/imap0_9d.pdf Signed-off-by: Alexander Graf <agraf(a)suse.de> Modified: trunk/openbios-devel/drivers/macio.c trunk/openbios-devel/drivers/pci.c trunk/openbios-devel/include/openbios/pci.h Modified: trunk/openbios-devel/drivers/macio.c ============================================================================== --- trunk/openbios-devel/drivers/macio.c Mon Feb 22 19:52:25 2010 (r678) +++ trunk/openbios-devel/drivers/macio.c Mon Feb 22 19:52:26 2010 (r679) @@ -18,6 +18,7 @@ #include "macio.h" #include "cuda.h" #include "escc.h" +#include "openbios/pci.h" #define OW_IO_NVRAM_SIZE 0x00020000 #define OW_IO_NVRAM_OFFSET 0x00060000 @@ -172,9 +173,13 @@ fword("finish-device"); if (is_newworld()) { + u32 *interrupt_map; + int len, i; + /* patch in interrupt parent */ dnode = find_dev(buf); - target_node = find_dev("/pci"); + + target_node = find_dev("/pci/mac-io"); set_int_property(target_node, "interrupt-parent", dnode); target_node = find_dev("/pci/mac-io/escc/ch-a"); @@ -182,6 +187,15 @@ target_node = find_dev("/pci/mac-io/escc/ch-b"); set_int_property(target_node, "interrupt-parent", dnode); + + target_node = find_dev("/pci"); + set_int_property(target_node, "interrupt-parent", dnode); + + interrupt_map = (u32 *)get_property(target_node, "interrupt-map", &len); + for (i = 0; i < 4; i++) { + interrupt_map[(i * 7) + PCI_INT_MAP_PIC_HANDLE] = (u32)dnode; + } + set_property(target_node, "interrupt-map", (char *)interrupt_map, len); } } Modified: trunk/openbios-devel/drivers/pci.c ============================================================================== --- trunk/openbios-devel/drivers/pci.c Mon Feb 22 19:52:25 2010 (r678) +++ trunk/openbios-devel/drivers/pci.c Mon Feb 22 19:52:26 2010 (r679) @@ -278,6 +278,47 @@ set_property(dev, "bus-range", (char *)props, 2 * sizeof(props[0])); } +static void pci_host_set_interrupt_map(const pci_config_t *config) +{ +/* XXX We currently have a hook in the MPIC init code to fill in its handle. + * If you want to have interrupt maps for your PCI host bus, add your + * architecture to the #if and make your bridge detect code fill in its + * handle too. + * + * It would be great if someone clever could come up with a more universal + * mechanism here. + */ +#if defined(CONFIG_PPC) + phandle_t dev = get_cur_dev(); + u32 props[7 * 4]; + int i; + +#if defined(CONFIG_PPC) + /* Oldworld macs do interrupt maps differently */ + if(!is_newworld()) + return; +#endif + + for (i = 0; i < (7*4); i+=7) { + props[i+PCI_INT_MAP_PCI0] = 0; + props[i+PCI_INT_MAP_PCI1] = 0; + props[i+PCI_INT_MAP_PCI2] = 0; + props[i+PCI_INT_MAP_PCI_INT] = (i / 7) + 1; // starts at PINA=1 + props[i+PCI_INT_MAP_PIC_HANDLE] = 0; // gets patched in later + props[i+PCI_INT_MAP_PIC_INT] = arch->irqs[i / 7]; + props[i+PCI_INT_MAP_PIC_POL] = 3; + } + set_property(dev, "interrupt-map", (char *)props, 7 * 4 * sizeof(props[0])); + + props[PCI_INT_MAP_PCI0] = 0; + props[PCI_INT_MAP_PCI1] = 0; + props[PCI_INT_MAP_PCI2] = 0; + props[PCI_INT_MAP_PCI_INT] = 0x7; + + set_property(dev, "interrupt-map-mask", (char *)props, 4 * sizeof(props[0])); +#endif +} + static void pci_host_set_reg(const pci_config_t *config) { phandle_t dev = get_cur_dev(); @@ -344,6 +385,7 @@ pci_host_set_reg(config); pci_host_set_ranges(config); pci_set_bus_range(config); + pci_host_set_interrupt_map(config); return 0; } Modified: trunk/openbios-devel/include/openbios/pci.h ============================================================================== --- trunk/openbios-devel/include/openbios/pci.h Mon Feb 22 19:52:25 2010 (r678) +++ trunk/openbios-devel/include/openbios/pci.h Mon Feb 22 19:52:26 2010 (r679) @@ -24,6 +24,16 @@ extern const pci_arch_t *arch; +/* Device tree offsets */ + +#define PCI_INT_MAP_PCI0 0 +#define PCI_INT_MAP_PCI1 1 +#define PCI_INT_MAP_PCI2 2 +#define PCI_INT_MAP_PCI_INT 3 +#define PCI_INT_MAP_PIC_HANDLE 4 +#define PCI_INT_MAP_PIC_INT 5 +#define PCI_INT_MAP_PIC_POL 6 + /* Device classes and subclasses */ #define PCI_BASE_CLASS_STORAGE 0x01

1 0

[commit] r678 - in trunk/openbios-devel: arch/ppc/qemu include/openbios
by repository service 23 Feb '10

23 Feb '10

Author: agraf Date: Mon Feb 22 19:52:25 2010 New Revision: 678 URL: http://tracker.coreboot.org/trac/openbios/changeset/678 Log: PPC: Get timebase and clock speed from fw_cfg When running on Qemu (TCG) time goes by at a constant frequency which accidently happens to match with Linux's fallback value. As soon as we go to KVM, the time base suddenly starts going as fast as the host CPU's time base. So we need to make sure Linux knows about it, so it can do its timekeeping properly. Signed-off-by: Alexander Graf <agraf(a)suse.de> Modified: trunk/openbios-devel/arch/ppc/qemu/init.c trunk/openbios-devel/include/openbios/fw_cfg.h Modified: trunk/openbios-devel/arch/ppc/qemu/init.c ============================================================================== --- trunk/openbios-devel/arch/ppc/qemu/init.c Mon Feb 22 19:52:23 2010 (r677) +++ trunk/openbios-devel/arch/ppc/qemu/init.c Mon Feb 22 19:52:25 2010 (r678) @@ -194,6 +194,16 @@ push_str("icache-block-size"); fword("property"); + PUSH(fw_cfg_read_i32(FW_CFG_PPC_TBFREQ)); + fword("encode-int"); + push_str("timebase-frequency"); + fword("property"); + + PUSH(fw_cfg_read_i32(FW_CFG_PPC_CPUFREQ)); + fword("encode-int"); + push_str("clock-frequency"); + fword("property"); + push_str("running"); fword("encode-string"); push_str("state"); Modified: trunk/openbios-devel/include/openbios/fw_cfg.h ============================================================================== --- trunk/openbios-devel/include/openbios/fw_cfg.h Mon Feb 22 19:52:23 2010 (r677) +++ trunk/openbios-devel/include/openbios/fw_cfg.h Mon Feb 22 19:52:25 2010 (r678) @@ -36,6 +36,12 @@ #define FW_CFG_ARCH_LOCAL 0x8000 #define FW_CFG_ENTRY_MASK ~(FW_CFG_WRITE_CHANNEL | FW_CFG_ARCH_LOCAL) +#define FW_CFG_PPC_WIDTH (FW_CFG_ARCH_LOCAL + 0x00) +#define FW_CFG_PPC_HEIGHT (FW_CFG_ARCH_LOCAL + 0x01) +#define FW_CFG_PPC_DEPTH (FW_CFG_ARCH_LOCAL + 0x02) +#define FW_CFG_PPC_TBFREQ (FW_CFG_ARCH_LOCAL + 0x03) +#define FW_CFG_PPC_CPUFREQ (FW_CFG_ARCH_LOCAL + 0x04) + #define FW_CFG_INVALID 0xffff #ifndef NO_QEMU_PROTOS

1 0

[commit] r677 - trunk/openbios-devel/arch/ppc/qemu
by repository service 23 Feb '10

23 Feb '10

Author: agraf Date: Mon Feb 22 19:52:23 2010 New Revision: 677 URL: http://tracker.coreboot.org/trac/openbios/changeset/677 Log: PPC: Increase instruction cache size A real 970FX has more icache, so we should reflect that. Signed-off-by: Alexander Graf <agraf(a)suse.de> Modified: trunk/openbios-devel/arch/ppc/qemu/init.c Modified: trunk/openbios-devel/arch/ppc/qemu/init.c ============================================================================== --- trunk/openbios-devel/arch/ppc/qemu/init.c Sun Feb 7 21:25:32 2010 (r676) +++ trunk/openbios-devel/arch/ppc/qemu/init.c Mon Feb 22 19:52:23 2010 (r677) @@ -416,7 +416,7 @@ { // XXX find out real values .iu_version = 0x003C0000, .name = "PowerPC,970FX", - .icache_size = 0x8000, + .icache_size = 0x10000, .dcache_size = 0x8000, .icache_sets = 0x80, .dcache_sets = 0x80,

1 0

[PATCH] Probe PCI buses recursively
by Blue Swirl 08 Feb '10

08 Feb '10

I tried to get PCI probing work in cases where the devices are behind multiple bridges. This patch works in many cases (especially current QEMU PPC/Sparc64 setup), but not if I make some arbitrary changes to the bridging setup. The mysterious "?active-package get-package-path" piece resolves the simple /pci/pci path (which would be identical to any other bridge found) to /pci@1fe0,0/pci@1,1. Is there a better way? Other comments? Signed-off-by: Blue Swirl <blauwirbel(a)gmail.com> --- drivers/pci.c | 72 +++++++++++++++++++++++++++++++++++++++++++------------- drivers/pci.h | 2 + 2 files changed, 57 insertions(+), 17 deletions(-) diff --git a/drivers/pci.c b/drivers/pci.c index 7d681f1..293509b 100644 --- a/drivers/pci.c +++ b/drivers/pci.c @@ -877,8 +877,15 @@ ob_pci_configure(pci_addr addr, pci_config_t *config, int num_regs, int rom_bar, pci_config_write16(addr, PCI_COMMAND, cmd); } -static void ob_scan_pci_bus(int bus, unsigned long *mem_base, - unsigned long *io_base, char **path) +static struct ob_buses{ + char *path; + pci_addr addr; +} ob_buses[256]; + +static int buses, found_host; + +static int ob_scan_pci_bus(int bus, unsigned long *mem_base, + unsigned long *io_base) { int devnum, fn, is_multi, vid, did; unsigned int htype; @@ -887,8 +894,9 @@ static void ob_scan_pci_bus(int bus, unsigned long *mem_base, const pci_dev_t *pci_dev; uint32_t ccode; uint8_t class, subclass, iface, rev; - int num_bars, rom_bar; + int num_bars, rom_bar, child_buses; + child_buses = 0; activate_device("/"); for (devnum = 0; devnum < 32; devnum++) { is_multi = 0; @@ -923,10 +931,10 @@ static void ob_scan_pci_bus(int bus, unsigned long *mem_base, if (pci_dev == NULL || pci_dev->name == NULL) snprintf(config.path, sizeof(config.path), - "%s/pci%x,%x", *path, vid, did); + "%s/pci%x,%x", ob_buses[bus].path, vid, did); else snprintf(config.path, sizeof(config.path), - "%s/%s", *path, pci_dev->name); + "%s/%s", ob_buses[bus].path, pci_dev->name); #ifdef CONFIG_DEBUG_PCI printk("%s - ", config.path); #endif @@ -957,37 +965,67 @@ static void ob_scan_pci_bus(int bus, unsigned long *mem_base, (subclass == PCI_SUBCLASS_BRIDGE_HOST || subclass == PCI_SUBCLASS_BRIDGE_PCI)) { /* host or bridge */ - free(*path); - *path = strdup(config.path); + fword("?active-package"); + fword("get-package-path"); + ob_buses[buses].path = pop_fstr_copy(); + ob_buses[buses].addr = addr; + pci_config_write8(addr, PCI_PRIMARY_BUS, bus); + pci_config_write8(addr, PCI_SUBORDINATE_BUS, + 255); + /* we assume the first device found is + the host */ + if (!found_host) { + pci_config_write8(addr, + PCI_SECONDARY_BUS, + buses + 1); + } else { + pci_config_write8(addr, + PCI_SECONDARY_BUS, + buses); + } + buses++; + child_buses++; + if (!found_host) { + found_host = 1; + } else { + int ret; + + ret = ob_scan_pci_bus(buses - 1, + mem_base, + io_base); + pci_config_write8(addr, + PCI_SUBORDINATE_BUS, + buses - 1 + ret); + child_buses += ret; + } } } } device_end(); + return child_buses; } +/* Recursively scan PCI bus, starting from host bus */ int ob_pci_init(void) { int bus; unsigned long mem_base, io_base; - char *path; #ifdef CONFIG_DEBUG_PCI printk("Initializing PCI devices...\n"); #endif - /* brute force bus scan */ - - /* Find all PCI bridges */ - mem_base = arch->mem_base; /* I/O ports under 0x400 are used by devices mapped at fixed location. */ io_base = arch->io_base + 0x400; - path = strdup(""); - for (bus = 0; bus<0x100; bus++) { - ob_scan_pci_bus(bus, &mem_base, &io_base, &path); - } - free(path); + + ob_buses[0].path = strdup(""); + ob_scan_pci_bus(0, &mem_base, &io_base); + + for (bus = 0; bus < buses; bus++) { + free(ob_buses[bus].path); + } return 0; } diff --git a/drivers/pci.h b/drivers/pci.h index 3ca71f3..fc8968f 100644 --- a/drivers/pci.h +++ b/drivers/pci.h @@ -34,7 +34,9 @@ #define PCI_HEADER_TYPE_NORMAL 0x00 #define PCI_HEADER_TYPE_BRIDGE 0x01 #define PCI_HEADER_TYPE_CARDBUS 0x02 +#define PCI_PRIMARY_BUS 0x18 #define PCI_SECONDARY_BUS 0x19 +#define PCI_SUBORDINATE_BUS 0x1a #define PCI_BASE_ADDR_0 0x10 #define PCI_BASE_ADDR_1 0x14 #define PCI_BASE_ADDR_2 0x18 -- 1.6.2.4

3 4

[commit] r676 - trunk/openbios-devel/drivers
by repository service 08 Feb '10

08 Feb '10

Author: blueswirl Date: Sun Feb 7 21:25:32 2010 New Revision: 676 URL: http://tracker.coreboot.org/trac/openbios/changeset/676 Log: Add "reg" property to kb_ps2, fix device_type Signed-off-by: Blue Swirl <blauwirbel(a)gmail.com> Modified: trunk/openbios-devel/drivers/pc_kbd.c Modified: trunk/openbios-devel/drivers/pc_kbd.c ============================================================================== --- trunk/openbios-devel/drivers/pc_kbd.c Sun Feb 7 21:25:30 2010 (r675) +++ trunk/openbios-devel/drivers/pc_kbd.c Sun Feb 7 21:25:32 2010 (r676) @@ -201,7 +201,7 @@ push_str(dev_name); fword("device-name"); - push_str("keyboard"); + push_str("serial"); fword("device-type"); PUSH(-1); @@ -209,6 +209,11 @@ push_str("keyboard"); fword("property"); + PUSH(0); + fword("encode-int"); + push_str("reg"); + fword("property"); + chosen = find_dev("/chosen"); push_str(nodebuff); fword("open-dev");

1 0

[commit] r675 - trunk/openbios-devel/drivers
by repository service 08 Feb '10

08 Feb '10

Author: blueswirl Date: Sun Feb 7 21:25:30 2010 New Revision: 675 URL: http://tracker.coreboot.org/trac/openbios/changeset/675 Log: Fix wrong device_type for non-PCI bridge devices, like EBUS Signed-off-by: Blue Swirl <blauwirbel(a)gmail.com> Modified: trunk/openbios-devel/drivers/pci_database.c Modified: trunk/openbios-devel/drivers/pci_database.c ============================================================================== --- trunk/openbios-devel/drivers/pci_database.c Wed Feb 3 18:54:14 2010 (r674) +++ trunk/openbios-devel/drivers/pci_database.c Sun Feb 7 21:25:30 2010 (r675) @@ -362,7 +362,7 @@ static const pci_subclass_t bridg_subclass[] = { { PCI_SUBCLASS_BRIDGE_HOST, "PCI host bridge", - NULL, hbrg_devices, NULL, + "pci", hbrg_devices, NULL, NULL, NULL, }, { @@ -382,7 +382,7 @@ }, { PCI_SUBCLASS_BRIDGE_PCI, "PCI-to-PCI bridge", - NULL, PCIbrg_devices, NULL, + "pci", PCIbrg_devices, NULL, NULL, NULL, }, { @@ -1040,7 +1040,7 @@ /* 0x05 */ { "memory controller", "memory-controller", mem_subclass, }, /* 0x06 */ - { "PCI bridge", "pci", bridg_subclass, }, + { "PCI bridge", NULL, bridg_subclass, }, /* 0x07 */ { "communication device", NULL, comm_subclass,}, /* 0x08 */

1 0